ST: Multi-Agent Decision Making

Catalog description

his course presents fundamental concepts and state-of-the-art research in sequential decision making that involves a single decision-making agent as the starting point and ultimately multiple agents that can cooperate or compete strategically. Topics covered include single-agent planning and reinforcement learning and their extensions to multi-agent games. Coursework includes mathematics and programming assignments that analyze and implement the presented single- and multi-agent decision making algorithms, as well as reading and critiquing research papers. Students can choose to do a project that is tailored to their interests and backgrounds. Recommended Background: Basic knowledge of linear algebra and probability theory, and experience with machine learning.